Using machine learning with supplemented NC code to predict machining energy

Manufacturing contributes a significant amount of value to the economy while consuming nearly one-third of the total energy produced within the nation. Computer-aided manufacturing tools arose to streamline the process of manufacturing parts, but they lack energy-conscious practices. With this gap, research has been done in the development of mechanistic and data-driven models to accurately predict the energy consumption of this process. However, validation carried out in the experimental methodology in many of the research models is ill-fit to represent their realistic counterparts. In this paper, a data-driven deep learning model is developed, which properly accounts for the complexities associated with CNC machining, as it compensates for variations in operations observed during CNC machining. Furthermore, this deep learning model makes predictions by processing supplemented NC programs sequentially. These programs include additional information regarding the material removal process. Four variants of the model are created to provide insights into the effects of supplementing the program with different material removal variables. The variables include depth of cut, width of cut, material removal rate, and the volume of material removed per numerically controlled instruction. The prediction capability of these four models are then compared using several statistical tests.